A METHODOLOGICAL APPROACH FOR IDENTIFYING ARCHAEOLOGICAL EVIDENCE OF FIRE RESULTING FROM HUMAN ACTIVITIES

The identification of unequivocal evidence of humanly-controlled fire is extremely problematic in archaeological contexts where obvious remains, such as charcoal and ash, have not been preserved. Until recently, the difficulty of identification had been magnified by the lack of an established methodological approach which addressed the problem. A methodological approach which can be used to identify and discriminate between archaeological evidence of fire resulting from natural processes and archaeological evidence of fire resulting from human activities is now available. The approach utilizes several parameters of investigation, including: (1) macroscopic examinations of fire feature surfaces and profiles, (2) analyses of magnetometer data, (3) magnetic susceptibility studies, (4) analyses of alternating field demagnetization characteristics, (5) studies of isothermal remanent magnetization (IRM) acquisition, (6) analyses of coercivity of remanence characteristics, and (7) palaeointensity studies. The methodological approach is illustrated by comparing the results from analyses of features and samples from grass fire experiment sites, isolated tree stump fire experiment sites, multiple-burn campfire experiment sites, modern forest fire sites, known archaeological hearths from Modoc Rock Shelter, Illinois, and known archaeological clay ovens from Tell Ashkelon, Israel, for each available parameter of investigation. Comparisons revealed that multiple-burn campfire sites exhibited basin-shaped three-dimensional configurations of oxidized sediments and diagnostic magnetic anomalies, while multiple-burn campfire samples displayed dramatic increases in magnetic susceptibility, highly stable to very highly stable responses to alternating field demagnetization, dramatic differences in IRM acquisition and coercivity of remanence responses, and diagnostic palaeointensity results, when compared with the responses of adjacent unfired control samples. Archaeological hearth features and samples, and samples from archaeological clay oven features, exhibited results which were broadly consistent with the results from multiple-burn campfire sites and samples. The results displayed by grass fire, isolated tree stump fire, and forest fire sites and samples were generally found to be mutually exclusive from the results exhibited by multiple-burn campfire and archaeological hearth and oven sites and samples for all parameters of investigation. The results demonstrate that the methodological approach can be used to unequivocally identify archaeological evidence of humanly-controlled fire resulting from multiple-burn campfires, fire hearths, ovens, and similar kinds of humanly-produced fires. © 1993 Academic Press, Inc.